Modern history


The Struggle over China’s Bond Markets

“If it doesn’t have access to a stable and sufficient source of capital, the China Development Bank will be unable to operate normally.”

Unnamed staff member, Treasury Department, China Development Bank

January 11, 20101

The combination of bank restructuring and the stock market’s collapse from mid-2001 catalyzed dynamic growth in China’s bond markets. This period began with the appointment of Zhou Xiaochuan to the governorship of the PBOC in early 2002. That year, a total of RMB933 billion (US$113 billion) in bonds, largely Chinese government bonds (CGB) and policy-bank bonds, was issued. By 2009 new issuance had nearly tripled, to RMB2.8 trillion (US$350 billion), and included a large chunk of corporate and bank bonds. As of year-end 2009, the total value of China’s outstanding stock of debt securities had reached RMB 17.5 trillion (US$2.6 trillion), with a mix of products that included government bonds (US$841.8 billion), PBOC notes (US$620.6 billion), bank bonds (US$747.1 billion) and a variety of corporate debt (US$354.1 billion) being traded between more than 9,000 institutional investors.

Many issuers struggled to get a piece of this market, none more significant than the China Development Bank (CDB). As the trends in Figure 4.2 illustrate, the CDB has begun to challenge for the dominant position, becoming, in effect, the country’s second Ministry of Finance. The bank’s RMB620 billion (US$912 billion) in issuance in 2009 was nearly on a par with the MOF’s RMB666.5 billion (excluding savings bonds) and represented nearly 30 percent of the total market. Equally significant, driven by the need to finance the stimulus package, Beijing at last recognized the legitimate funding needs of local governments and allowed certain poor provinces to issue bonds. In addition, all levels of local government made aggressive use of the bond markets, raising RMB423 billion (US$62 billion) via their own enterprises on top of massive levels of bank borrowing.

Far surpassing the CDB and the localities were the PBOC’s efforts to sterilize the creation of new RMB generated by China’s huge inflows of foreign currency. From 2003, as China’s trade surplus began to widen and foreign investors flocked to invest, the PBOC began to issue ever-increasing amounts of short-term notes (and sometimes long-term notes, as we saw in Chapter 3) to control the domestic money supply. This use of a market-based tool to manage the macro-economy was a first in China, but pressures on the PBOC grew to the point that its institutional rival, the MOF, was able to step in and “help out.” A complex series of transactions relating to the establishment of China’s second sovereign-wealth fund revealed a triumphant MOF in control of the very lynchpin of the domestic financial system, once again bringing the story back to the bank dividend policies described earlier.


The rapid growth in the China Development Bank’s funding and lending activities coincides with the ascension of the new Party and government leadership in 2003 and the start of a continuing debate about China’s economic development model. During this time, the CDB became the darling of those supporting a return to both a more state-planned economy domestically and a natural-resource-based foreign policy internationally. But understanding the CDB’s position is complicated by the fact that, on the one hand, it represented a challenge to the prevailing banking model sponsored by Zhou while, on the other, it depended on the PBOC for approval of its annual bond-issuance plan. The dramatic increase in its bond issuance during this period may be the result of the PBOC’s antipathy toward the MOF; but the MOF also had its own tactics.

Chen Yuan, the bank’s very ambitious founding chairman, positioned the CDB deliberately as an alternative model to the Big 4 banks. The Big 4, under Zhou Xiaochuan’s reform program, followed a path modeled after their international counterparts, including the deliberate introduction of international banks as strategic investors. As we saw earlier, Chen Yuan was opposed to what he referred to as “that American stuff.” Instead, he proposed to “develop around our own needs and build our own banking system” which, he said, “must provide the capital to meet the needs of our high-growth economy, resolve the various financial bottlenecks of our enterprises and provide a channel for capital for various types of enterprise.”2

The bank’s investment projects, once included in the national budget, are now independent of it; the CDB can, to a certain extent, determine on its own “commercial” principles what projects to invest in and what not. Nonetheless, its projects are state projects and its obligations are state obligations. The CDB, unlike the Big 4 banks, was established as a ministerial-level entity with quasi-sovereign status reporting directly to the State Council. It is a typical example of an organization, not an institution, built around one man, the son of a powerful revolutionary-era personage. Chen’s father, Chen Yun, was the planner whose famous “Bird Cage” theory provided the ideological foundation for the Special Economic Zones in the 1980s. Political conservatives were able to accept the idea of foreign investment as long as it was “caged” inside these special zones. This powerful political concept has now morphed and provides the inspiration for the distinction between “inside” and “outside” the system. Unless there are bounds imposed by a determined Party leader, as was the case during Zhu Rongji’s era,3 “princelings” such as Chen Yuan can drive the political and economic process in ways contrary to the national interest, as we will see further in the next chapter. In Chen’s case, the goal has long been to add both an investment bank and a securities company to the CDB portfolio and become China’s first universal bank (and this in spite of his avowed aversion to “American stuff”). If the CDB can be a universal bank, is it any wonder that the Big 4 banks have reacted defensively by wanting to acquire the licenses held by their respective AMCs?

If it had succeeded, the strategy Chen marked out for the CDB would have marked a return of China’s banking system to the pre-reform era and the People’s Construction Bank of China (PCBC). Essentially a division of the MOF, the PCBC provided exactly the same kinds of long-term capital services for the economy as Chen’s CDB. The difference, however, is that the CDB possesses a modern corporate veneer and polished public-relations expertise, as evidenced by its website on which Chen’s old-fashioned sloganeering is pushed into the background. That, however, is not the most important difference. The PCBC was funded by the national budget and it channeled, on behalf of the MOF, the disbursement of interest-free investment funds to SOEs and special infrastructure projects contained in the state plan. But the CDB does not rely on the national budget for funding.

This fact and Chen’s own ambitions created a trap for the CDB. As a policy bank, the CDB funds itself through debt issuance in the markets, and China’s bond markets are fully reliant on the commercial banks and the PBOC for support. Some 72 percent of the CDB’s funding comes from those very banks Chen holds in such low esteem. As Figure 5.1 illustrates, beginning in 2005, CDB bond issues began to grow rapidly. This growth was approved by the PBOC, whose own position was being challenged by the MOF.

FIGURE 5.1 MOF vs. policy bank bond issuance, 2001–2009


Source: China Bond

Note: 2007 MOF issuance excludes the CIC Special Bond of RMB 1.55 trillion; 2009 MOF issuance excludes RMB200 billion local government bonds.

Since banks are the source of all funds, if this situation had continued, market saturation could have been reached and, from that point, the CDB would have begun to squeeze out the MOF, as Figure 5.1 indicates. Can one man fight the MOF as well as the Big 4 Banks? The answer is “No” and this is where ambition led Chen Yuan astray.

In addition to pursuing greater business scale, Chen appears to have envied the superficial modernity of the commercial banks. This led him to seek to create China’s first universal bank and to become publicly listed. He was clearly encouraged to pursue this goal and on December 11, 2008, the CDB became a joint-stock corporation, the first step in preparing for an IPO. But why in the world would the State Council seek to make a policy bank—which had been designed to make and hold non-commercial investments in state-designated infrastructure projects—into a commercial entity and then to list it? Chen’s argument was that “commercialization” would not change the bank’s strategy as a development bank:

The lesson we can take is that a first-class bank should not take even the very best Western bank as a standard. We should have an objective international standard . . . expressed as having high-quality assets, the trust of market investors, and an objective, fair and deep understanding of society’s needs and to work to resolve those social needs in order to receive society’s approval.4

The CDB model, with its emphasis on social-justice issues, threatens the last 15 years of profit-oriented banking reform in China and has already erased the policy-bank reforms of 1994. While Chen’s words and vision may have resonated among the country’s people and top leadership in the wake of the global financial crisis, they found no such resonance in the bond market.5 The reason for this is simple: the banking regulator had been less than precise in defining the CDB’s new status. After its incorporation, is the CDB still a semi-sovereign policy bank or is it now a commercial bank? From this definition flows the valuation of its outstanding bonds, as well as the pricing for future bond issues. Market pricing will differ depending on the answer and there is as yet no clear answer. It is not just about pricing: if the CDB is a commercial entity, other banks and insurance companies will only be able to invest up to a regulation-imposed limit and the CDB’s days of cheaply funded expansion will rapidly come to an end.

The uncertainty as to the CDB’s exact status is what accounts for its bonds being more actively traded than those of the MOF. Uncertainty is risk and the risk created by Chen’s ambition is by no means small for his investors, the commercial banks. They currently hold an estimated RMB3.2 trillion (US$460 billion) of his old bonds on their balance sheets and even a small 0.5 percent drop in value for them would mean mark-to-market losses of RMB16 billion (US$2.3 billion). Even though they hold CDB debt as investments, the loss of value on such a scale might ultimately require their international auditors to recommend loss provisions and a hit to their income.

In early 2010 as the party reflected on the massive lending binge of 2009, the accepted line became: “We know these projects do not produce cash flow today, but they will prove very useful to China’s development later on.” This describes perfectly the function of lending by policy banks. On the other hand, policy loans in commercial banks are, by definition, NPLs. This was recognized in 1994 when the CDB was established and the Big 4 banks embarked on commercialization. It is now beyond irony that just as China’s hard-built commercial banks have turned themselves into policy banks, the China Development Bank is striding in the opposite direction.

Without question, this market-based outcome has marked a defeat for Chen Yuan’s ambitions for the CDB and represents a major victory for the MOF, which in all probability encouraged his listing ambitions. As discussed below, when China Investment Corporation (CIC) acquired Central Huijin in 2007, it acquired a full 100 percent interest in the CDB. The tables were now turned on Chen Yuan. In the greatest of political ironies, the CDB has now returned to its roots—as a mere sub-department of the MOF—but at what cost to the system?

The People’s Bank of China and the National Development and Reform Commission

In contrast to the CDB’s aspirations and the MOF’s bloodless revenge, the now discredited market-oriented model espoused by reformers is far less eloquently described.6 It involved the development of direct, market-based, enterprise-financing capabilities based on the decisions of enterprise management. In other words, corporations were to be given a choice between banks and debt markets. Not only that, they were to take responsibility for their decisions for both shareholders and bond investors; in short, the full international capital-markets model. To create this possibility, in 2005, in the midst of collapsed domestic stock markets, the PBOC leveraged a regulatory loop hole defining “corporate bonds”7 as those with maturities above one year. It used this definition to create a short-term debt product, commercial paper (CP; duanqi rongziquan image), that quickly became the debt product of choice among SOEs.

In 1993, the PBOC had ceded the corporate-debt product to the State Planning Commission (SPC) primarily because issuers would not take responsibility for repayment of bonds on maturity. This created huge difficulties at the time and largely caused the product to be terminated. But in 2005, corporate bonds were a hot product again due to both bank reform and the weak equity markets. Unfortunately, the “enterprise bond” (qiyezhai image) market belonged to the SPC’s grandchild, the National Development and Reform Commission (NDRC), with underwriting done only by securities companies regulated by the CSRC. As the volume of issuance in the years up to 2005 illustrate (see Figure 5.2), neither agency gave much thought to developing this product. From the NDRC’s perspective, a bond was an afterthought. The few projects contained in its planning documents were funded by the national or local budgets or the banks and it could see no need to develop the bond product. From the CSRC’s perspective, bonds represented a zero-sum game with equity products, and the regulator’s avenue to achievement was not fixed-income products or markets.

FIGURE 5.2 Issue volume by product type,1992–2009


Source: PBOC, Financial Stability Report, various

Note: 2007 CGB issuance excludes the RMB1.55 trillion Special Bond

Zhou Xiaochuan provided a detailed analysis of the corporate market’s resultant inadequacies in his famous October 2005 speech excerpted at the start of this chapter.8 He rightly pointed out that the root cause of the market’s failure to develop was found in the command-economy mentality of the “early days of the transition when the economy was more planned than market-driven.” This comment, historically couched as it was, pointed straight at the NDRC, but the fact is that previous central bank administrations had also done little to promote the bond markets, leaving them to the MOF.

With the support of the Party’s “Nine Articles,” which explicitly called for the development of bond markets, the PBOC drove through this “one year and above” loop hole and created a CP market out of thin air. In 2005, its first year, more than RMB142 billion (US$17 billion) in CP was issued by presumably capital-starved SOEs. This amount tripled in 2008, with growth being driven by a unique ease of issuance: no regulatory approvals were required, only registration. PBOC reformers modeled this process after that used in the US wherein issuers are required to have a credit rating (this takes about three weeks in China), an underwriter (banks, which are not regulated by the CSRC), a prospectus, and a filing with the PBOC. To further ease the government out of any role in the market, in September 2007, the PBOC sponsored the establishment of an industry association—the National Association of Financial Markets Institutional Investors (NAFMII)—to manage things. In contrast to the opaqueness of China’s equity markets, the universe of debt issuers, their financials, approval documents and prospectuses are available for all to see online on the China Bond website.

NAFMII is registered as a non-profit, non-government organization authorized by the PBOC to advise on the development of the debt-capital market, to sponsor new policies and regulations, and to review debt issues. When establishing NAFMII, the PBOC was astute enough to create a governing board including the Who’s Who of China’s banking industry. In its brief existence, the agency has become the regulator in charge of the most rapidly growing segments of the inter-bank debt market, including local-currency risk-management products. Its scope of authority would, of course, exclude the NDRC’s enterprise bonds (qiyezhai image), as well as financial bonds and subordinated bank debt which, given their direct impact on the sensitive banking sector, remain directly with the PBOC.

If the commercial paper ploy did not upset the NDRC, the PBOC’s next move did. In April 2008, the PBOC, working through NAFMII, created a three–to–five-year medium-term note (zhongqi piaoju image). Unlike bonds, which are issued once and remain outstanding until redemption or maturity, MTNs are issued like CP as part of a “program” that allows the issuer, depending on his funding needs, to issue more or less of the securities within a certain overall limit. Perhaps a bit sarcastically, the NAFMII called these securities “non-financial enterprise financing instruments” (feijinrong qiye rongzi gongju image) in order to clearly demarcate them from the NDRC’s “enterprise” bonds and the CSRC’s “company” bonds. MTNs, like CP, only require registration with NAFMII.

The NDRC, however, did not find the wordplay funny and sought to stop the PBOC and its MTNs by claiming control over the notes which, after all, had tenors of more than one year. The State Council accepted the case, delaying the product’s debut for four months. Later in the year, however, a consensus developed that more debt in the right hands would bolster the swooning stock markets and MTNs were given the go-ahead. In just three months, enterprises raised RMB174 billion (US$26 billion), in new capital and, in 2009, the market grew explosively. By year-end, some RMB608 billion(US$ 89 billion) in new MTNs had been issued. Together with CP, these new instruments accounted for 22 percent of the total capital raised in the fixed-income markets in 2009.

With the defeat of its approach to financial reform in 2005, the PBOC could only push the development of new products in its own space as part of the infrastructure for the future. As Figure 4.10 shows, CP and MTNs with their shorter maturities brought in new non-state investors, mutual funds and foreign banks. For the first time in China’s bond markets, such investors played a significant role, accounting for 30 percent of short-term corporate-debt holdings. Such small victories can, over time, add up to something important when circumstances change.


The PBOC’s product innovations provided a solution to financing problems for all sorts of Chinese corporations and not just those commonly thought of as SOEs. It has been 15 years since the last serious reform of China’s system of taxation created a clear split between those taxes belonging to the center and those to the localities. Since that time, SOE reform, the closures of hundreds of failed local financial institutions and the centralization of bank management have greatly reduced the financial resources available to local governments. The shortfall between revenues and expenditures has widened significantly.

In its 2009 budget report to the National People’s Congress (NPC), the MOF confirmed that, overall, local governments ran major fiscal deficits. The report stated that total local revenues amounted to RMB5.9 trillion (US$865 billion), of which RMB2.89 trillion (US$423 billion) derived from tax-transfer payments from the central government. Set against this, local-government expenditures were RMB6.13 trillion (US$900 billion). The life of a provincial governor or city mayor is dominated by a scramble to raise capital in support of local development and new jobs. Previously, SOE reform—selling off poorly performing SOEs outright or listing the shares of good ones—and attracting large amounts of foreign investment had shown the way forward for the most commercially sophisticated provinces. But there are few provinces that share the commercial attractions of China’s coastal areas. There are only so many SOEs that can be publicly listed, even on supportive Chinese exchanges. After the Asian Financial Crisis, the poorly performing SOEs had been privatized or closed entirely to preserve local resources. As a result, to increase their budgets and service their debt, local governments rely on cash flow from projects and land auctions, which reportedly contribute over one-third of local extra-budgetary revenue.

The global financial crisis of 2009 posed the greatest challenge yet to localities: Beijing’s RMB4 trillion (US$486 billion) stimulus package required local governments to identify projects and come up with financing for two-thirds of project spending. For some time before the crisis, local governments had been leveraging their utilities, roads, construction brigades and asset-management bureaus by incorporating them into limited-liability companies. Under this legal guise, they could borrow money from banks and, taking advantage of bond-market reform, issue debt. According to the CBRC, by June 2009, there were 8,221 fund-raising platforms operating at provincial, regional, county and municipal government levels, of which the majority (4,907) were owned by county governments. Many of these entities had been established simply to take advantage of the government’s free-for-all lending boom. After all, if they could come up with the capital to meet Beijing’s demands, why not raise more to finance their own economic incentive programs? It is common wisdom in China that once the window of opportunity is open, it is open for only the briefest of moments and the wise person will grab all that is possible of whatever opportunity is on offer. It is also common wisdom that when the Party takes the responsibility, the regulators will sit silently on the sidelines. Thus, in 2009, conditions were perfect for local governments to do all in their power to raise funds, with little possibility of their being blamed for financial excess.

“Financing platforms”

In those easy days of 2009, local governments and their “financing platforms” (rongzi pingtai image) had almost-unprecedented access to credit. After a long discussion in 2008, Beijing decided that local governments would be officially permitted to run fiscal deficits. The symbol of this new thinking was the RMB200 billion (US$30 billion) in bonds issued in early 2009 by the MOF as agent for the localities. Even more important, however, locally incorporated investment companies and utilities were allowed to issue NDRC-approved enterprise bonds (qiyezhai image) and to these were added the new short-term PBOC securities, CP and MTNs. With the window wide open, local Party secretaries rapidly expanded their fund-raising platforms beyond simple “Municipal (or County) Investment Companies” to the incorporation of various entities such as water, highway, and energy utilities. China’s muni bond market was born.

Among the many new issuers in the inter-bank market were some 140 local-government incorporated entities (see Table 5.1). With such names as Shanghai Municipal Construction Investment and Development Co. Ltd., Wuhan Waterworks Group Co. Ltd. and Nanjing Public Utility Holdings Co. Ltd., these entities are similar to municipal-bond issuers in the US, but with one difference. In addition to issuing long-term bonds to match the maturity of long-term investment projects, these entities also eagerly issued short-term commercial paper and MTNs. In fact, it seems they have issued every sort of debt security for which they could obtain approval.

TABLE 5.1 Local “financing platform” debt issuance, June 30, 2009

Source: Wind Information; bonds include CP, MTNs and enterprise bonds; issuers do not include local manufacturing SOEs



In 2009, the amount of capital raised in the bond markets by these provincial, municipal and county entities totaled nearly RMB650 billion (US$95 billion), accounting for over 50 percent of enterprise bonds issued and 48 percent of total CP and MTN issues. The overall explosion in CP and MTN issuance in 2009 is explained by the lack of complex approval procedures and of the requirement of a bank guarantee; issues had only to be registered with NAFMII. Making it even more attractive, MTN underwriting fees and interest expenses were lower than for the NDRC’s bonds or even bank loans. Truly, 2009 was a bonanza for some local governments. Looked at carefully, the geographical distribution of these local issuers is quite limited; fully 66 percent of local government issuers and 76 percent of all money raised came from China’s richest locations: Greater Shanghai, Beijing and Guangdong. What can explain why Zhejiang, China’s richest provincial economy, has 19 local-government issuers down to even the county level, while Henan, China’s most populous province, only four? In discussion with market participants, it seems the answer is simply that money begets money.

So the other 8,000 less-fortunate localities depended, as has been the case historically, on their partnerships with local bank branches for debt financing. How do they borrow if their resources are so constrained? Figure 5.3 shows that one of the ways local governments capitalize their financing platforms is by contributing land and tax subsidies. The land, in turn, may be used as collateral for a bond issue or a bank loan.

FIGURE 5.3 Local-government funding alternatives


Source: Based on FinanceAsia, June 2009: 32

The more valuable the land, the stronger is the platform’s capacity for borrowing. The mortgaged land might be used as part of a large development for houses, office space or shopping centers. The stronger a local economy, the greater the potential profit of such a development and the greater the interest other investors may have in participating in the equity of the platform through wealth-management products developed by trust companies and sold to the banks’ high-net-worth customers (see Figure 5.4).

FIGURE 5.4 Trust-based financing for local financing platforms


In China’s many poorer localities, these sorts of opportunities do not exist. They simply cannot meet the minimum standards of the bond markets, so they borrow from banks, as they have always done. To do so, they cut corners. For example, Figure 5.5illustrates how a local government may borrow from another local government and use the money to inject as capital in its financing platform. Once the capital is registered and the company is established, the local government takes back the money and repays the other local government. The financing platform exists, has nominal registered capital and business licenses and is fully able to borrow money from other banks.

FIGURE 5.5 The temporary debt-based equity capitalization of a local financing platform


PBOC and CBRC surveys found such local platforms had borrowed some RMB6 trillion (US$880 billion) by the end of September 2009, with nearly 90 percent of stimulus projects tied to bank loans.9 These same surveys also noted that these loans amounted to 240 percent of local government revenues and, in 13 of China’s 29 provinces and autonomous zones, liabilities exceeded total fiscal revenues. These local borrowings were found by the CBRC to account for 14 percent of total lending in 2009 and, for some banks, as much as 40 percent of total new credit issued. By year-end 2009, Beijing was publicly admitting to RMB7.8 trillion (US$1.14 trillion) in outstanding local-government bank debt.

If this widely reported figure is accurate, then local governments and their agencies have borrowed the equivalent of 23 percent of the country’s annual GDP. Analysts at China International Capital Corporation (CICC) put current debt levels at RMB7.2 trillion (US$1.1 trillion), peaking at RMB9.8 trillion (US$1.4 trillion) in 2011. But the report provides an even greater service when it states: “If the financing chain for the platforms is not broken, they will be able to dissolve potential credit risks in the course of current economic development trends.”10 In other words, as long as banks continue to lend, there will be no repayment problems. As will be discussed in this book’s conclusion, rolling outstanding debt when it matures—that is, re-issuing new debt to repay the old—is a principal characteristic of China’s financial system.

Provincial government semi-sovereign debt

As noted earlier, in March 2009, the MOF announced the issue of RMB200 billion (US$30 billion) of local-government bonds. These were rolled out rapidly, before a regulatory framework could be developed or the philosophy behind their credit ratings could be thought out. The thinking behind this seemed to be that as provincial governments are equivalent to ministries in the Chinese bureaucratic hierarchy, and thus represent the state, they would attract similar ratings. But most people know full well that Qinghai is different from Shanghai.

However, a spokesman for the MOF explained that although the debt would be issued in the name of the provincial governments and approved as a part of their budgets, the MOF would act as their agent. More importantly, given this, the coupon on, and the risk of, the bonds would approach that of sovereign debt. In essence, the spokesman was attempting to sell the idea that provincial issues carried risk equivalent to the country itself. While this may be true in theory, in practice, it is not and many market participants did not accept the idea. Moreover, this was not the picture presented in the MOF’s own rules governing local debt issues. These stated explicitly that if the locality could not make repayment, the debt would be rolled forward over a period of one to five years, with the original principal amount being repaid in installments the local government could afford. In short, the original three-year bond might, in fact, have a tenor of eight years or longer.

Clearly, this was not central government debt. Despite widespread questioning in the market, the bonds were priced closely to MOF bonds of a similar maturity by the friendly primary-dealer syndicate. What might have happened to prices in the secondary market is unknown because there has been no secondary market. The fact is, these bonds have disappeared onto the balance sheets of the MOF’s primary-dealer group, that is, China’s banks.


If the inter-bank market somewhat resembles a pyramid scheme, it nonetheless plays an extremely important role in the PBOC’s effort to manage inflation through control of the money supply. In 2002, a policy disagreement blew up over how to cool inflation that was threatening at that point. This disagreement evolved over four years into a struggle over how to manage the inflationary impulse caused by China’s huge trade surpluses. Driven by enthusiasm over China’s joining the WTO, fixed-asset investment exploded from 2002, increasing on an annual basis by 31 percent, the highest level seen since Zhu Rongji’s heyday began in 1993 (see Figure 5.6). Memories of mid-1990s double-digit inflation caused the PBOC to issue short-term notes continuously into the market for the first time in China’s post-1949 banking history. From an initial issue equivalent to US$26 billion in 2002, this rose over the following years until 2007, when the PBOC soaked up nearly US$600 billion from the banks. The central bank also adjusted upward bank-deposit reserve ratios nine times and raised interest rates five times. These aggressive measures were effective temporarily, but by 2007, the explosion in foreign reserves and the consequent creation of new renminbi posed an almost insurmountable challenge.

FIGURE 5.6 Investment, FX reserves and money supply, FY2001–2008


Source: PBOC, Financial Stability Report, various

Others also remembered 1993 and how Zhu Rongji had aggressively intervened in the economy using administrative orders to close off all channels to liquidity. Zhu simply shut off all bank lending, closing down the economy for almost three years until inflation that had peaked at over 20 percent in 1995 was finally beaten. The group favoring administrative intervention in 2002 argued that the economy was not overheating, only specific industrial sectors were and this could be dealt with using specific policy means. Their argument carried the day and bank credit to those sectors was cut off. By 2004, both efforts had reduced the growth of investment and M2 just in time to begin dealing with the flood of US dollars pouring into the country from China’s booming trade surplus.

It was at this point in July 2005 that the PBOC succeeded in convincing the government to delink the RMB exchange rate from the US dollar and allow it to gradually appreciate. It was unfortunate that the predictability of the RMB’s 20 percent appreciation led to huge inflows of hot money and even greater liquidity in the domestic markets, not to mention an explosion in the stock index and high-end real-estate speculation. That both sides could claim success in this earlier anti-inflation campaign complicated things significantly later on after the outbreak of the global financial crisis in September 2008, when market forces fell dramatically from political favor. Despite the PBOC’s aggressive efforts to manage this flood of RMB, the data in Figure 5.7 suggest that the effectiveness of short-term notes began to decline after 2007. The growth in the money supply as measured by M2 had been stable at 16 percent, but it accelerated to 19 percent in 2008 and then 29 percent in 2009. For its part, the policy of appreciating the currency was canceled as soon as export growth turned negative in late 2008.

FIGURE 5.7 PBOC note issuance vs. growth in money supply (M2), 2001–2009


Source: PBOC Financial Stability Report, various; China Bond

This is the macro economic background to the political struggle over China’s financial framework that had been continuing since 2003. The struggle came to center on the establishment in 2007 of China Investment Corporation (CIC). It is ironic to consider that this sovereign-wealth fund, established to invest the country’s FX reserves overseas, was in fact used to dramatically restructure China’s own financial system. CIC is not the country’s first sovereign-wealth fund. SAFE Investment Co. Ltd., a Hong Kong subsidiary of the State Administration of Foreign Exchange (SAFE), has been actively managing a portion of China’s foreign-exchange reserves since 1997. So why establish a second fund? SAFE Investment is owned by the PBOC; CIC is owned by the MOF. Why would the MOF want to encroach on foreign exchange, which has clearly always been the legitimate turf of the PBOC? The answer seems to be that since the PBOC took over outright control of two of the four major state banks from the MOF, the MOF had the right to seek their recovery. In the end, the establishment of CIC is less about a sovereign-wealth fund than a battle over bureaucratic territory. The outcome of this particular round, moreover, is very clear: CIC is now the very lynchpin of the country’s domestic financial system.

RMB sterilization and CIC

The story of CIC’s capitalization shows that all institutional arrangements in China are impermanent; everything can be changed as a result of circumstances and the balance of political power. All institutions are in play, even the oldest and most important. The case of CIC also shows the extent to which China’s financial markets have been distorted by the pressures created by the country’s tremendous foreign-reserve imbalance. This distortion now extends beyond the domestic capital markets, both debt and equity, to the financial institutions that provide their foundation and beyond to the international equity markets and investors.

Since it was designed to invest reserves offshore, it might have been expected that CIC would receive its capital directly from foreign reserves, as had the state banks. China Construction Bank, Bank of China, Industrial and Commercial Bank of China and Agricultural Bank of China had each been at least partially capitalized from the foreign reserves by way of a PBOC-established entity called Central SAFE Investment, known commonly as Huijin. In 2007, a surging money supply threatened a major asset bubble and the debate about how to handle this—whether through monetary tools or outright administrative measures—became mixed up with the MOF/PBOC rivalry. The MOF claimed that the PBOC’s management of reserves produced too low a return; from this it was a quick jump to the MOF asking for its own opportunity and then to a discussion of how to capitalize CIC. In the end, the Party agreed to allow the MOF a chance; after all, in 2007, there were plenty of reserves to be managed. But there was no direct infusion of capital into CIC as had been the case with the banks. Instead, there was yet another MOF Special Treasury Bond.

This Special Bond was approved by the State Council in early 2007 and its size was a mammoth RMB1.55 trillion (about US$200 billion with 10- and 15-year tenors), as shown in Figure 5.8. Not only had the MOF accused it of poor reserve management, the PBOC also stood blamed for monetary growth that threatened an outbreak of inflation. The path these bonds took tells the tale of the PBOC’s political weakness. The MOF sold the bonds to the PBOC in eight separate issues through the hands of Agricultural Bank of China. Direct dealings between the two had been prohibited by law since 1994 when the Central Bank Law was enacted; prior to this, the central bank had too often been forced to finance the state deficit directly. But this bond was not for deficit financing. The PBOC, for its part, bought the bonds from ABC and then, given their below-market interest coupons, forced them into the market, which consisted of the banks. This issue, therefore, drained a huge amount of liquidity from the banking system, an amount double what the PBOC had been able to achieve through its own short-term notes. This approach also relieved the PBOC of adding to a growing interest burden on its notes.

FIGURE 5.8 Step 1: MOF issues Special Bonds and drains market liquidity


While this seemed like an innovative idea, nothing is without cost, as shown in Step 2 (see Figure 5.9). Considering the monetary objective, the MOF had done the PBOC a favor and the transaction could have stopped there. CIC could have been funded through a separate transaction with Huijin using foreign reserves, if the PBOC had been willing to go along. The MOF, however, was out to extract its final pound of flesh and used the RMB acquired from the bond issue to buy US$200 billion from the PBOC/SAFE, again via the services of ABC. The MOF then used these funds to capitalize CIC. Aside from its economic objectives, the institutional effect of this transaction was the restoration of the financial system to the pre-2003 status quo and a further weakening of the PBOC and the market-reform camp. But there was more.

FIGURE 5.9 Step 2: MOF buys US$ from the PBOC to capitalize CIC


After the money had changed hands, CIC for all intents belonged to the MOF. Although it reported directly to the State Council, its top senior management came from the MOF system. This was not necessarily a loss for the reform camp, since CIC was also staffed at senior levels with officials historically associated with market reform. But there was an awkward technical problem arising from the MOF’s arrangement: how would interest be paid on its underlying Special Bonds? The cost involved amounted to around US$10 billion annually. The surprising answer was that CIC would pick up the interest. As the head of CIC dryly commented, the burden was about RMB300 million for each day CIC was open for business. How would CIC, a newly established entity not meant to be a short-term investor, have the immediate cash flow to cover such a huge obligation? The solution to this problem by itself has put an end to further hope of bank reform.

Careful calculation, however, had been given to this solution; it reveals how in 2007 the Party desired to organize China’s financial system and goes to the heart of the PBOC’s loss of institutional clout. Even before CIC received its new capital, the US$200 billion had been budgeted and spent, and only one-third of it related to its advertised mission as a sovereign-wealth fund. The other two-thirds, some US$134 billion, was to be spent on, first, a planned recapitalization of ABC, the CDB and other banks and financial institutions and, second, the outright acquisition of Central Huijin from the PBOC. In one stroke, CIC became China’s financial SASAC.

One might ask why a sovereign-wealth fund would want to own or invest in domestic financial institutions already owned outright by the government: the money would just be going around in circles. But this is exactly what happens “inside the system.” The attractive professional face its management presents internationally is belied by the reality that CIC is, at best, only a part-time sovereign-wealth fund. Its most important role is to serve as the lynchpin of China’s financial system. That system has been restored to one inspired by the old Soviet model centered on the MOF and with a weak central bank.

The MOF is the obvious winner in this domestic game and it is a purely status-quo power. Its victory ultimately has significant negative implications for continued bank reform and can clearly be seen on referring back to Table 3.3, which shows the pre- and post-IPO controlling shareholders of China’s major banks. From the very start of bank reform in 1998, the fundamental point of contention between the MOF and the PBOC has been over which entity owns the state banks on behalf of the state. Once reform entered its critical stages in 2003, the plans of Zhou Xiaochuan’s group of reformers began to affect the economic and political interests of the MOF. When Huijin recapitalized CCB and BOC, it did so in a way that established direct economic ownership and, at the same time, used the MOF’s equity interest to write off problem loans; no longer did these two banks “belong” to the MOF empire. In the years after Zhou’s political defeat in 2005, the MOF won back control, starting with the ICBC and continuing through ABC, at least in part because of the PBOC’s difficulties managing its primary responsibilities: inflation and the currency. But this was hardly the only reason.

A contributing part of the PBOC’s political weakness during the crucial year of 2005 was the terminal illness suffered by Vice Premier Huang Ju, who was in charge of the financial sector. In early 2005, Huang stepped aside for treatment and the Premier assumed his portfolio. Consensus politics resumed and the consensus was that Zhou Xiaochuan had overreached. Reform is not about consensus. It was an easy decision that rocked no boats to allow the MOF to retain as equity its 1998 capital contribution in ICBC. So Huijin, after injecting US$15 billion, received only 50 percent of ICBC and the MOF’s 1998 contribution remained. The bureaucratic pendulum had begun its backward swing. By late 2007, with CIC’s outright acquisition of Huijin, the status quo had been fully restored.

The argument in favor of this acquisition was simple: CIC was responsible for the interest on the Special Bonds. Acquiring the banks gave it access to their dividend stream.11 Since it all belonged to the state anyway, what difference did it make? The fact is, it did make a difference and not just to the bureaucracies involved. If CIC were to acquire Huijin, then SAFE would want its original investment back. The US$67 billion price represented the original net-asset value of its investments in all three banks and a collection of bankrupt securities companies.12 Since it would be simply a transfer of state-owned assets between state agencies, by government rules, there need be no premium paid. It was just a matter of accounting and the money going from one pocket to the other. The change in “owners,” however, would have a huge impact on reform moving forward.

For CIC, the acquisition worked out very well. According to its first full-year financial report for FY2008, CIC carries at a market value of US$171 billion what it acquired from the PBOC for only US$67 billion. This increased in 2009 to over US$200 billion but, of course, included investments other than Huijin at that point. These investments by Huijin allowed CIC to claim a profit in its first year and so helped deter criticism of its controversial (and loss-making) investments in Blackstone and Morgan Stanley. But there was one small detail that had not been properly considered: three of the Big 4 banks were now internationally listed: the Party was no longer just playing “inside the system.”

CIC squeezes its banks

Beyond its mark-to-market profit on its bank investment portfolio, CIC also relied on the banks for the cash flow to make its interest payments on the MOF bonds, not to mention to pay dividends to MOF that helped it meet its obligations on the IOUs given to ICBC and ABC. The Huijin arrangement designed by the PBOC team placed CIC in a position to receive dividends paid by the banks (see Chapter 7 for further details). This rich source of cash had originally been designed to help offset the unrecoverable loans the PBOC had made to the AMCs in support of bank restructuring. The Huijin arrangement acted as a form of taxation that would, over time, have reduced the PBOC’s credit losses and strengthened its balance sheet.

When CIC acquired Huijin in late 2007, it acquired direct economic control over China’s major banks via their boards of directors and a decisive vote in the matter of their dividends. There would be no intervening levels of ownership, management, and powerful Party secretaries to muddy the issue (as was the case in the SASAC’s state-owned enterprises). Huijin was controlled by the PBOC which, in turn, is controlled by the Finance Leadership Group at the very top of the Party hierarchy. In sum, now MOF was in a position to recommend, if not decide, how much was to be received by . . . itself.

This takes the story full circle back to the bank IPOs. The dividend payout ratio of around 50 percent for all three banks is not necessarily excessive by international standards for banks that are growing at a stable rate and in a normal business environment where bad loans and securities losses are not material. This, however, does not describe the situation faced by the Chinese banks. These banks drive national economic growth by increased lending typically at 20 percent a year and in certain years, such as 2009, much more. But from 2008, Huijin’s new duty would be to pay interest on CIC’s bond obligations to the MOF as well as help the MOF make good on its IOUs. Since the banks are publicly listed and audited by international accounting firms, cash dividends paid by them are transparent to all. To some extent, they might serve as a reliable indicator of how the government thinks about its banks.

In the early days of restructuring, as NPLs were being spun off to the AMCs and capital rebuilt, asset growth was tightly controlled (see Figure 3.7 for the years 2001–2005) and capital ratios were rapidly bolstered. After their respective listings, however, lending, profits and dividends for all three banks grew rapidly, particularly after CIC acquired Central Huijin in 2007. From that point, total dividends immediately increased to a level sufficient to cover CIC’s interest obligations, leaving plenty left over to reduce the outstanding restructuring IOU due to ICBC from RMB143 billion (US$18 billion) to RMB62 billion (US$9 billion) (see Figure 5.10). Of course, to the extent that CIC and its banks were responsible for making such payments, the national budget was freed of the obligation.

FIGURE 5.10 Big 4 bank IPOs, cash dividends paid and CIC, 2004–2009


Source: Huijin; bank annual audited financial statements

These financial arrangements raise questions about the future path of China’s bank reform. Given the experience of 2009, there is no question but that banks have reverted to their former business model as the Party’s financial utility. But is it really possible that dividend policy is being set to meet the MOF’s own parochial needs? The appearance certainly suggests that the listed banks have become cash cows subsidizing the MOF’s efforts which, among other things, are aimed at sidetracking the institutional influence of the PBOC. Worse yet, it is outrageous that the full amount of cash dividends paid during this period has been funded by the IPOs of state banks (as discussed in Chapter 2). From a very simplistic point of view, international and domestic investors handed over US$42 billion in new capital to the banks and indirectly to the MOF, yet received in those years less than US$8 billion in dividends. Beyond that, is it possible that under pressure to maintain dividends, bank managements might easily be encouraged to increase lending? With fixed spreads over the cost of funding, more lending assures more earnings, higher dividends, better stock price and higher rankings on the Fortune Global 500. Then came the economic stimulus package, which provided all the excuse needed to do just that.

Not even a year later, in early 2010, however, the combination of 50 percent dividend payouts and binge lending have created huge challenges for the banks. Most challenged of all must be Bank of China, whose loan portfolio grew nearly 43 percent in 2009 while the other major banks hit levels over 20 percent. Given the high dividend payouts and asset growth, it is hardly surprising that the banks rapidly grew out of their capital base, with BOC and ICBC rapidly approaching capital ratios close to pre-IPO levels (see Table 5.2).

TABLE 5.2 Trends in core capital-adequacy ratio, 2004–1H 2010

Source: bank annual and interim audited reports


From that point, there was much government hand-wringing as to how bank capital could be increased. In early 2010, each of the banks announced record 2009 earnings and improved NPL ratios . . . and one after the other, each has announced plans to raise for a second time that US$40 billion chunk of capital they had raised from their IPOs and then paid out to the state (see Table 2.3). Of course, the state would be required to disgorge capital as well if it desired to maintain its shareholding. So it was not surprising when rumors emerged that Huijin, the direct majority shareholder of the major banks, was seeking approval for a large capital injection of up to US$50 billion to match its share of bank capital and maintain its equity position.13 Even more interesting, CIC had requested an additional US$200 billion from the MOF. Both requests were subsequently cut back significantly, Huijin to an RMB190 billion (US$28 billion) bond issue and CIC to US$100 billion.

The dividends, the excessive lending and the scramble for new capital can all be ascribed, at least in part, to the MOF’s acquisition of banking assets from the PBOC. Had China Investment Corporation been capitalized directly from China’s foreign-exchange reserves, it could have remained a pure sovereign-wealth fund and the MOF would have had its own counterpoint to SAFE Investments. Had there really been the need to sterilize such a massive amount of RMB, the Special Bond could have been issued separately. But the MOF combined the two and the resulting structure twisted the heart of China’s financial system into this awkward bureaucratic and economic position.

What to do with Huijin is perhaps the biggest topic on the agenda of the Fourth National Finance Work Conference in mid-2010. This is part of a much broader power grab by the MOF, which hopes to use Huijin as the basis of a “Financial SASAC” that would become, among other things, the Super Regulator for China’s entire financial sector, replacing “one bank and three commissions.”14 Even if this were to happen and Huijin were to be freed of CIC, the arrangement with regard to the Special Bond would likely remain. Then there is the question of which state entity would pay CIC the US$67 billion it is nominally worth and where the money would come from. The point of this is that Huijin and its banks continue to be the object of a bureaucratic ping-pong game domestically that increasingly exposes the internationally listed banks to the valuation judgment of international investors precisely at the time that the government has actively desired to cut back foreign influence.


It is well recognized that China’s currency policy of fixing the RMB exchange rate against the US dollar greatly limits flexibility in interest rates. This by itself means that real fixed-income markets cannot readily develop. There is another dimension to this problem. China’s banks depend on Party-guaranteed profitability created by mandated minimum spreads between deposits and lending rates. The profit generated here from corporate borrowers subsidizes their “investment” in sub-market-priced government securities. This can work only so long as they operate in a protected domestic oligopoly well insulated from outside pressures. Foreign banks exist in China only to provide the suggestion of an open market. With profits guaranteed, banks have never had to be creative in competing for customer support. Nor have they had to worry about new capital or problem loans: these are the Party’s problems, not those of bank management. So when the Party calls for development of the bond market, the banks follow, even though bonds are little more than disguised loans. The corporate-bond market stops there: there is no secondary market. But the fixed-income market is more than just corporate bonds.

In recent years, the flood of US dollars from a large trade surplus and inflows of hot money, the consequent creation of new RMB, the need to sterilize that RMB to prevent inflation and asset bubbles have combined to distort the very institutions on which the financial system is built. When in 2007 the MOF argued that PBOC notes were insufficient to offset excess RMB, the ensuing political solution attached itself to the wholly unrelated establishment of CIC. It was argued that CIC’s capitalization solved two major problems: temporarily controlling money creation and putting to use a large portion of the country’s foreign reserves. This was a clever ad hoc solution that became complicated by CIC’s acquisition of Huijin.

Leveraging Huijin’s bank investments to pay interest on the MOF’s bonds may have seemed a good idea at the start; it appears that the Party mistakenly believes its own advertising about its banks being rich and strong. But it linked the stresses of China’s domestic financial markets directly to the international financial markets. This has created an economic and political exposure contrary to the fundamental interests of the “system.” An unconvertible currency, fixed exchange and interest rates and the need for strong bank lending to drive GDP growth create inevitable and predictable demand for huge amounts of new bank capital that, in turn, depends on international and domestic capital markets. With over US$70 billion in new capital to be raised, these markets will, in the end, be demanding and price sensitive, even if many friends of China internationally and domestically stepped up to make Agricultural Bank of China’s US$20 billion IPO an apparent success as it was.

An economic stimulus package that in retrospect appears to have been excessive gave banks a free option to expand their lending. But stimulus or not, this is what the Party’s banks do in any circumstances, as history has shown. With mandated loan spreads, RMB10 trillion (US$1.5 trillion) in new loans grew bank earnings dramatically. It is important to note as well that banks are happy to lend to local governments—at the behest, of course, of the local Party secretary—directly or through bonds. Can they go bankrupt? Are they lesser credit risks than SOEs? The interim announcements of the Big 4 banks in 2010 have been full of record profits and very high loan-loss reserves and, given rapid loan-portfolio growth, inevitably declining NPL ratios well below two percent. It is all a matter of simple mathematics and has nothing to do with strong management performance or value creation. There will be record dividend payouts and further improvements in their Fortune 500 rankings. But the lending explosion rapidly depleted bank capital. The first decade of the twenty-first century now appears to have ended, just as each of the last three decades of the twentieth century did, with China’s major banks in desperate need of massive recapitalization.

This marks the completion of one full cycle of China’s money machine; it has taken 10 years. But the fault lines, created playing by the rules “inside the system” while pretending to abide by international standards and regulatory requirements, have begun to be clear. The second cycle can now be reliably mapped out and illustrates why true reform of the system is unlikely. Mandated minimum loan-to-deposit spreads sustain bank profitability thereby guaranteeing that dividends can be paid out to investors, namely Huijin. Huijin, in turn, must meet the demands of CIC, which must meet the demands of the MOF Special Bond. In the cases of ICBC and ABC, too, the MOF must make repayments on its special IOU arrangements. Even if Huijin were separated from CIC, each year cash would flow up from the banks to the MOF and then from the MOF back to the banks. The banks will expand lending to borrowers to drive high GDP-growth numbers and generate greater profit as long as China’s export and non-state sector remains weak.

How, then, can the Party allow the banks to be disintermediated by capital markets or real outside competition? Protectionist measures, controlled exchange rates and fixed lending spreads ensure the Party’s control and the stability of the system, and virtually guarantee that the banks must raise new capital every few years to prime the cycle. Viewed from the outside, bank profits reassure retail depositors that their banks are sound and their deposits safe. International investors support bank shares since they are seen as proxies of a bank-driven GDP number. The banks use household deposits and new equity capital to fund new loans to drive GDP and to support the conceit that is China’s debt-capital market, which sustains the appearance of overall convergence toward a Western-style market system.

Instead of removing the risk burden from the banks, China’s backward bond markets create new risk. Making up around 30 percent of the total assets of the Big 4 banks, these “investment” portfolios enjoy negative interest spreads, leaving the banks exposed to significant market risk. In order to offset this weight, banks will inevitably lend more and increase credit risk. More asset bubbles, stock-market booms and problem loans are the inevitable product of this arrangement. The tools to deal with these problems, the AMCs, the MOF’s IOUs and the PBOC’s credit support, already exist. As has been shown with the first generation of bad loans, these measures have contained the problem and pushed the inevitable off into the future onto the agenda of the next Party leadership group and out of the memory of international observers. The cycles and the pressures that are building up “inside the system” can continue for a very long time. Where is the catalyst that will disrupt it? Even if the Emperor is ultimately seen to be naked, he is still the Emperor.


1 The Economic Observer image, January 11, 2010: 1.

2 The Economic Observer image, July 20, 2009: 41.

3 Zhu thwarted Chen’s first attempt in 1995 to establish an investment bank in favor of Wang Qishan’s joint venture with Morgan Stanley, CICC.

4 The Economic Observer image, July 20, 2008: 41.

5 The Economic Observer image, January 11, 2010: 1.

6 See Li Liming, “Liangnian zhongguo jinrong shengtai gaibianle image, (In two years China’s financial environment has changed),” The Economic Observer image, August 29, 2005: 10. This failed effort at reform in 2005 was picked up again in the major article by Yang Kaisheng, ICBC’s CEO, in early 2010, “Wending woguo shangye yinhang ziben chongzu shuiping de jidian sikao image (Several thoughts on stabilizing the capital adequacy levels of our country’s commercial banks,” 21st Century Business Herald 21image, April 13, 2010: 10.

7 It is confusing to translate “corporate bonds,” as there are two types: one controlled by the NDRC and traded in the inter-bank market (qiyezhai image), and one controlled by the CSRC and traded on the stock exchanges (gongsizhai image). Zhou’s loophole related to the NDRC regulations.

8 Zhou Xiaochuan, “Learn lessons from the past for the benefit of future endeavor,” Speech at the China Bond Market Development Summit, Beijing, October 20, 2005,

9 See Fang Huilei, Zhang Man, Yu Jing and Zhang Yuzhe, “Scary View from China’s Financing Platforms,”Caixin Magazine online, February 5, 2010; and Victor Shih, “Big rock-candy mountain,” China Economic Quarterly 14(2), June 2010.

10 21st Century Business Herald 21image, April 12, 2010: 6.

11 There is a second interesting attraction to using a bond and interest payments to channel money to the MOF: payment would not require any involvement of CIC’s board of directors, it would simply be business as usual. As the cash flow came in as dividends from its subsidiary banks, CIC’s CFO would simply pay interest when due to the MOF; no formal board decisions or minutes would need to be made. Contrast this with the SASAC.

12 Huijin paid US$22.5 billion each for BOC and CCB and US$15 billion for the ICBC, and US$7 billion for interests in a variety of securities companies. It carries these investments on its books at this same value, notwithstanding that the banks have all been listed and have a market value far above this number.

13 First rumored in November 2009 and then confirmed in April 2010. See Bloomberg, November 11, 2009; and Caixin image, April 23, 2010; Asian Investor, April 1, 2010.

14 This refers to the PBOC, the CSRC, the CBRC, and the China Insurance Regulatory Commission. For the background, see The Economic Observer image, July 12, 2010: 2.

If you find an error please notify us in the comments. Thank you!